Comparison of Artificial Neural Network Algorithm for Water Quality Prediction of River Ganga

摘要

The development of any region depends greatly on the availability of appropriate water supplies. The quality of water can be judged based on a variety of parameters among which the most important is the temperature. In this study, Artificial Neural Network algorithms, Lavenberg Marquardt (LM) and Gradient Descent Adaptive (GDA) have been used to predict the quality of water. Using the data of temperature for the year 2008 to 12, researchers have measured Biochemical Oxygen Demand (BOD) and Dissolved Oxygen (DO) along River Ganga. Both the algorithms, mentioned above, have been compared for their performance. The results show that the algorithm LM gives a better performance as compared to that of GDA. Hence, simulated values for the desired locations at which measured data are unavailable can be efficiently provided by a trained ANN Model.